Neural Adaptive Distributed Formation Control of Nonlinear Multi-UAVs With Unmodeled Dynamics

Area of Science:

• Robotics
• Control Systems
• Artificial Intelligence

Background:

• Quadrotor unmanned aerial vehicles (UAVs) face challenges in formation control due to unmodeled dynamics and external disturbances.
• Existing control methods often struggle to address the complexities of multi-UAV systems operating in uncertain environments.

Purpose of the Study:

• To develop a robust neural adaptive distributed formation control strategy for multiple quadrotor UAVs.
• To address unmodeled dynamics and external disturbances for stable and accurate formation flight.

Main Methods:

• A virtual position controller based on backstepping was designed for the position subsystem.
• A distributed formation scheme utilizing local UAV information and inter-UAV communication was proposed.
• A neural adaptive sliding mode controller (SMC) was developed to compensate for compound uncertainties.

Main Results:

• The proposed controller effectively compensated for nonlinearities, unmodeled dynamics, and external disturbances.
• The distributed formation scheme enabled UAVs to update their states based on neighboring information.
• Lyapunov theory confirmed that tracking errors converge to a small neighborhood of zero.

Conclusions:

• The developed neural adaptive distributed formation control scheme is effective for multi-UAV systems.
• The strategy ensures robust performance and accurate formation tracking even with system uncertainties.
• Simulation results validate the proposed approach for practical applications in quadrotor UAVs.